Privacy Policy

Questions you ask on our blog

Our blog has a box where you can ask a question about a post. You do not need an account, and we do not ask for your name or email.

• What we keep. The text of your question, and a one-way hash of your network address. We do not store the address itself — the hash exists only to stop one person flooding the box, and cannot be turned back into an address.
• Where it goes. Your question, together with extracts from posts already published on our blog, is sent to an AI provider so an answer can be drafted. Nothing about any VidTraffic account is included, and the provider does not use it to train models.
• What is published. Nothing, unless we decide to publish it. An answer you are shown immediately is marked as unreviewed and is visible only to you. If we later publish it, your question appears on the post alongside the answer — so do not put anything personal in it.
• How long we keep it. Questions we have not published are deleted after 90 days. Published ones stay, because at that point they are part of the post.

Account emails

We use Resend to deliver account emails, including a welcome message when you sign up. We share your email address and, when the message uses it, your first name, together with the email content. We do not include your YouTube videos, channel data, or analytics in these emails. These account emails do not subscribe you to a marketing mailing list. When we send a message to a group of account holders, we choose who receives it from account information, such as whether onboarding is complete, and never from your YouTube activity or analytics.

Pending emails and delivery records are retained in VidTraffic for up to 30 days. Message content and recipient information are removed from our queue when sending finishes. Deleting your account removes your pending emails and local delivery records. Resend processes delivery data under its own retention rules; deleting your VidTraffic account does not recall an email already sent or immediately erase the provider's delivery records. We do not use open or click tracking for these account emails.

Information we collect

• Google account information. When you sign in with Google, we receive your name, email address, and profile picture.
• YouTube data (read-only). With your permission, we access your YouTube channel and video information and your YouTube Analytics (such as views, watch time, subscriber changes, traffic sources, the search phrases viewers typed to find your channel, and how far into each video people watch) to show you insights inside the app. We never post, edit, or delete anything on your behalf.
• Comments on your videos. When you ask for them, we fetch the comments left on your recent videos so you can see what your audience is asking. We store the text and the like and reply counts only. We do not store who wrote a comment — the display name and channel of the person who left it arrive with it and are discarded, because answering “what is my audience asking” does not require knowing who asked.
• Content you submit. Titles, descriptions, thumbnail ideas, and images you provide so the app can generate and score suggestions.

How we use your information

• To provide the core features: generating, improving, and scoring your content.
• To display your channel analytics and competitor and trend insights.
• To suggest what to make next on your plan's Ideas tab, counted in our own code from the search phrases viewers typed, the questions in your comments, the channels you track and your Trends page. No AI is used for it. The comments themselves are shown only to you, as examples, and are never copied into your plan.
• To keep you signed in and operate and secure the service.

We do not sell your personal information, and we do not use your Google or YouTube data for advertising.

YouTube API Services

VidTraffic uses YouTube API Services. By using the app, you also agree to the YouTube Terms of Service and acknowledge the Google Privacy Policy. You can revoke VidTraffic's access to your data at any time via the Google security settings page.

Limited use of Google user data

VidTraffic's use and transfer of information received from Google APIs to any other app will adhere to the Google API Services User Data Policy, including the Limited Use requirements. We only use Google user data to provide and improve the features you request, and we do not transfer it to others except as needed to operate the service, to comply with the law, or with your consent.

Our staff tools do not show your videos, titles, descriptions, analytics or channel details. The one exception: a team member with the right role can look up the ID of the YouTube channel you connected, and a link to its public page, to support your account or to investigate abuse. Each lookup is recorded with who made it and when. It is not used for marketing or advertising, and it disappears with the rest of your YouTube data when you disconnect, revoke access, delete your account, or after 30 days.

Third-party processing

To generate and score content, the text you submit is processed by our AI provider, currently Google's Gemini on a paid tier, solely to return results to you. Our AI providers act as processors and are not permitted to use your data to train their models or to retain it under our agreements.

If you have connected a YouTube channel, what we send may also include signals measured from that channel — for example your best- and worst-performing titles, the keywords that recur in them, your typical view counts and retention, how much of your traffic comes from search, and the search phrases people actually used to find you — so that what you get back sounds like your channel rather than a generic one. This is Google user data, and it is handled under the Limited Use rules above: it is sent only to deliver the result you asked for, only to providers contractually barred from training on it or retaining it, and never for advertising, never sold, and never sent to any recipient you did not ask us to send it to. It is never used to build a profile of you for any purpose other than producing your own results.

The same applies to the written review on My Channel: the measured facts on that page — views, watch time, subscribers, click-through, traffic sources, audience splits, and the titles of the videos that earned the views — are sent to the AI provider, under the same rules, so it can tell you what is working and what to do next. The review is kept with your analytics and is deleted with them.

On My Channel you can also ask why viewers left one of your published videos. The moments where that video's retention curve falls or rises are picked by our own code from your YouTube Analytics. When you press it, we send Google's Gemini, on a paid tier that does not train on or retain it, short clips of that video taken from its public YouTube link at those moments, the times of those moments and whether viewers leave or come back there — never your view counts or retention figures — and the video's title and the start of its description, so it can say what is on screen and being said at each moment and suggest one change. It works only for your own videos and only when the video is public; for an unlisted, private or processing video nothing is sent. It runs only when this deployment has AI on channel data switched on. What comes back is kept with those moments for 30 days, deleted when you disconnect, and deleted when you delete your account.

Your chat conversations are stored and treated as YouTube data: a reply can quote your own figures, so conversations are kept for 30 days, deleted when you disconnect, and deleted when you delete your account.

Your plan is in two parts. An idea itself — its title, the note you write, the script you write — is yours: it survives a disconnect and is deleted only when you delete your account. The one exception is a title that quotes your YouTube data (a search phrase viewers typed, another channel's video title, or a topic counted from comments or other channels' titles): 30 days after you add the idea, or when you disconnect, it is replaced with a general one, such as “Make a video for a search phrase”, unless you have renamed it. The research an idea is built from is split by who wrote each piece. What we derived from your YouTube data (the finding an idea came from, a video and its figures, a search phrase people typed) is YouTube data, kept for 30 days and deleted when you disconnect. What you added yourself (a link, a paragraph, an image from your library) is yours and stays with the idea. Each piece says which it is.

The studio keeps what you make in it. Each time you use one of its tools we store what you typed and what came back — the titles and their scores, the descriptions, and the images, which go to your image library — against the video you are working on, so you can go back to any of them. That is your work, not YouTube data: nothing derived from your channel's figures is stored with it, it survives a disconnect, and it is deleted when you delete that video or your account. To make an image we send your words, any photos you chose to include and, when you mark part of an image to change, that image and the area you marked to Google's image model, on a paid tier that does not use them to improve its products; the marked area is used for that one change and not kept. When this deployment has AI on channel data switched on, Google's image model may also be given the audience and the thumbnail colours we derived from your channel, so the picture suits it, under the same rules. Where a deployment makes images with another service instead (Cloudflare Workers AI or Pollinations), that service is sent only the words for the picture and the styling you set yourself — never your photos and nothing derived from your channel. The studio needs an account.

The studio can also work from the video itself, if you give it one. Your browser reads the file on your device and takes out its sound and a small picture every 30 seconds; the video file itself is never uploaded. The sound goes straight from your browser to Google's Gemini API, and the pictures go with your request through our server, on a paid tier that does not use them to improve its products. They are used once, to read the video: the sound is deleted at Google as soon as it has been read, or the attempt fails, and anything an interrupted upload leaves behind is deleted within hours; the pictures are not stored at all. If you give a link to a public YouTube video instead, Gemini watches that video directly. What we keep is a written account of the video — what it covers, its chapters and their times, the names in it and the moments that would make a thumbnail — with that studio video, so its title, description and thumbnail can be made from it without reading it again. That account is your work, not YouTube data: it is deleted when you remove the video from the studio, delete that studio video, or delete your account. This needs you to be signed in, and it runs only when this deployment has that switched on.

The overlay maker's prompt box sends the request you type — “make it yellow, and faster” — with the overlay's settings and their current colours, sizes and choices, to Google's Gemini on a paid tier that does not use them to improve its products, so it can change those settings for you. The overlay's words are not sent, and nothing from your channel is. Nothing from it is kept beyond the count of requests you have made today.

When you ask the assistant about an idea or ask it to draft a script, the idea's note, research and script are sent to the AI provider together with your channel's facts, under the same rules as the written read: only to produce your own draft, only to a provider that does not train on or retain it, and only when this deployment has that switched on. Your open plan is also sent with any question you ask, so the assistant does not propose again something you already said you would do.

To answer a question, the assistant can also look up what VidTraffic already holds for you: your plan, your Settings profile, the written read of your channel, the competitors you track, what your Trends page shows, and your credits and plan — whether credits apply to your account yet, what is left and when it resets, what each tool costs, what you have used, and what each plan includes. What it looks up goes to the AI provider with your question, under the same rules: only to answer you, only to Google's Gemini on a paid tier that does not train on or retain it, and anything that comes from your YouTube data only when this deployment has that switched on. Your credits and plan are our own record, not YouTube data, and only yours are ever sent; where the assistant cannot look things up, a short summary of them — whether credits apply to you yet, what is left, when it resets and what each tool costs — goes with each question instead, so it can answer what things cost. To answer about trends it may bring YouTube's public most-popular chart for your country and category up to date, as opening Trends does, which uses one unit of your daily YouTube lookups; it never runs the niche scan, and otherwise fetches nothing new from YouTube. It can propose changes to your plan or to your Settings, but it never makes them itself: a proposal is a card in its reply, and nothing changes unless you press it. What it looked up is not stored; the reply you see, with any card in it, is kept with the conversation like any other reply.

The assistant can also use the studio's text tools when you ask: write titles, each with its score, write a description for a title, or score a title with its description and thumbnail idea. These are the studio's own tools, and they send what the studio sends — your words, your Settings profile and, if you name one of your studio videos, its written account — to the same AI provider, under the same rules. Writing titles or a description uses one of your daily generations, as it does in the studio; scoring uses none. When you name one of your studio videos, what it wrote is kept with that video like any other studio run. It does not make thumbnails: those are made in the studio.

So that an answer stays consistent with what you settled before, the earlier turns of a conversation, and short passages from your earlier conversations, may be sent back to the AI provider with a new question, under the same rules. With AI on channel data switched off, only your own questions are sent back, never the replies, which can quote your figures.

Disconnecting your channel stops this and deletes the signals we derived from it. Signed out — which means only the prompt box on our home page — or with no channel connected, generation works only from the context you type yourself and no Google data is involved.

Cookies and analytics

We use Google Analytics to understand which pages and features people find useful. It sets cookies in your browser and sends Google information about your visit — the pages you opened, roughly where you are, and what kind of device you used.

We also use Google Tag Manager, a Google service that loads the measurement tags our team chooses. Loading it sends Google a request from your browser, and the tags it loads collect the same kind of visit information described above, plus the names of steps you take in the app — for example that you started connecting YouTube or finished setting up your profile.

Three limits on both, enforced in the code rather than promised here:

• Nothing loads until you say yes. The first time you arrive we ask. If you decline, or simply do not answer, neither Google script is ever requested at all — not loaded-and-restricted, not requested.
• None of your YouTube data is sent to them. They run on our public pages and in the app, but they never receive your channel name, your video titles, your analytics or anything else we read from YouTube. Page addresses in the app carry none of it, the step names we send are fixed labels, and Tag Manager is locked so that nothing it loads can read what is on the page — no custom code, no tags from other companies, and no tags that record what you click. When you are signed in they also receive whether you are signed in and have connected YouTube, and a random identifier for your account — not your name, your email, your Google account or your channel — so visits can be counted as the same person across days and devices. It is cleared when you sign out.
• Never in our staff console. Neither runs on the internal tools our team uses.

You can change your answer at any time by clearing this site’s data in your browser, which makes us ask again on your next visit. Declining does not limit anything on the site.

Separately from this, the app stores a small number of items in your browser that are not analytics and are not optional: your sign-in session, your theme, and your answer to the question above. The overlay maker also keeps the overlay you are working on there, with any picture you add to it — a logo, or a screenshot of your video to aim by — and the brand colours you choose, so a reload does not lose them. Overlays are drawn and exported in your browser: none of this is sent to us or to any AI provider. If you put your channel's name in an overlay it is kept there like any words you type; your channel's avatar is never kept, and is fetched from your account again when an overlay shows it. Smart timing reads a video's retention curve from YouTube when you pick the video, shows you when to place an overlay, and keeps neither. They stay on your device and are not used to track you across other sites.

Connecting your own agent

You can give a program you choose — an agent — permission to read your channel data through us, so it can work with your own numbers on your behalf. You start this yourself, you can see what you have connected, and you can disconnect it at any time; access stops on the next request rather than at some later expiry.

An agent you connect can read only your own channel's data. It cannot reach anybody else's, and there is nothing you can grant it that would let it.

Once data reaches an agent you have connected, what that agent and the provider behind it do with it is governed by their terms, not ours — we cannot see it and cannot control it. So connect only agents you trust, in the same way you would think about any application you give account access to. We record every connection you make and every request an agent sends, so you can always find out what was read and by which agent.

Disconnecting your YouTube channel disconnects every agent along with it.

The VidTraffic browser extension

The extension does nothing on YouTube pages until you choose Continue in its first-run card. When its panel is open, it reads the page's address and links to find which videos and channels are there, and in YouTube Studio it reads what you type into a video's title, description and tags. What it reads stays in your browser. It sends us only the following:

• If you choose Look up numbers. While the panel is open, the IDs of the video or channel you are viewing, and of the videos listed on that page, sent with your sign-in. These IDs reveal what you are looking at. We use them only to fetch those videos' and channels' public statistics from YouTube's official API. We do not store which IDs came from you, and our service logs for these requests do not record your account. The statistics are kept, not linked to anyone, for no more than 30 days. We never send IDs from your History, Watch later, Liked videos, your own playlists or your Subscriptions feed. On a search page we send the results' IDs, not the words you typed.
• Things you ask for, when you press them. Saving a video or channel to your plan, or tracking a channel as a competitor, sends its ID. Asking for titles in your style on a video you are watching sends that video's ID: we read its title from YouTube ourselves and send it to our AI provider to write and score the titles. Scoring your draft or asking for a suggested title or description in YouTube Studio sends what you typed there — your title, description and tags — and nothing before you press. These AI requests go to Google's Gemini on a paid tier, under the same rules as the rest of VidTraffic described above, and use your allowances as they do in the app.
• Your own channel. On your own videos the panel shows numbers from your connected channel, such as click-through rate, retention and the searches that found the video, fetched with the access you gave us for the app.
• If you choose Share how you use the extension. Which parts of the extension you use and how often, linked to your VidTraffic account, never what you watch or search.

You can sign the extension out from its options page, or from Connected agents in the app, at any time. Disconnecting YouTube or deleting your account signs it out and deletes your account-linked data as described in this policy. The extension does not run in incognito windows, and it contains no advertising or analytics code.

Our use of information received from Google APIs will adhere to the Chrome Web Store User Data Policy, including the Limited Use requirements.

Payments

When you buy a plan or credits, the checkout is run by Paddle, our payment provider and the Merchant of Record for every purchase: Paddle (Paddle.com Market Ltd, or Paddle.com Inc. for buyers in the United States) is the seller of record, takes the payment, adds and pays any VAT or sales tax, and issues your invoice.

We send Paddle your account email address, so its checkout is filled in and your receipt reaches you, our internal ID for your account, and the plan and billing interval you chose — nothing else, and never any of your YouTube data. Your card or other payment details, your billing address and any tax number you give it go to Paddle directly; we never see or store them. Paddle processes them as an independent controller under its own privacy policy, and keeps its records of your purchases for as long as the law requires a seller to, even after you delete your account with us.

From Paddle we keep only what it takes to give your account the plan you paid for: Paddle's identifiers for your customer record, subscription and payments, the plan and interval, the subscription's status and its dates. No card details, no address and no amounts. Manage subscription opens Paddle's customer portal, where you can cancel, change your payment method and download invoices.

Data storage and security

Your data is stored on our server and protected with industry-standard measures, including encrypted connections (HTTPS). The access and refresh tokens that maintain your connection to YouTube are encrypted at rest and are never shared with anyone.

Data retention and deletion

• YouTube data. Channel, video, analytics, and comment data we fetch from YouTube is cached only to keep the app fast. Any of it that has not been refreshed within 30 days is automatically deleted, in line with the YouTube API Services Terms. Some of it — the day-by-day view chart and the retention curve for a single video — is fetched fresh each time you open it and never stored at all. When you ask why viewers left a video, the moments picked from its curve, with the share still watching at each, and what was said about them are kept for at most 30 days.
• If you revoke access. When you disconnect VidTraffic — either in the app or from your Google account — we revoke our access token and delete the YouTube data we hold for you, including the comments, search phrases and reads of why viewers left described above.
• Account data. Your account details and the content you created are kept while your account is open, and deleted when you delete it.
• Usage analytics. We record which pages you open and which features you use — the names of actions from a fixed list, never the content you type. Your video titles, descriptions, search terms and channel context are not part of it and cannot be, because the only values we accept are short labels from a predefined set. We store a first-party identifier on this domain to join your visits together, and it is never shared with anyone. This system is ours and runs everywhere you are signed in; Google Analytics is a separate thing that runs only if you accept it — see Cookies and analytics above. Kept for 90 days, deleted when you delete your account, and you can turn it off entirely in Settings. We also honour Do Not Track and Global Privacy Control without being asked.
• Credits. We keep a record of the credits on your account: what your plan gives you each month, and each time a feature uses them — the feature's name, how many credits and when, with internal reference numbers. It never includes what you typed or made. It is how your balance and your Usage are shown to you. If you press "Notify me" on a plan, we keep that you asked and for which plan. Both are deleted when you delete your account, and neither is sent to AI providers or used for advertising.
• Payments. If you have bought a plan or credits, our record of it: Paddle's identifiers, the plan and interval, the subscription's status and its dates. When you delete your account we first ask Paddle to cancel any subscription that is still running, at once, so you are not charged again, and then delete this record. Paddle keeps its own records of past payments, as a seller must. None of it is sent to AI providers or used for advertising.
• Abuse prevention. The prompt box on our home page works before you sign in, so we cap how many tries a single visitor can make in a day — otherwise anyone could run up our costs. Every other AI feature needs an account. To count them we store a one-way, salted hash of your network address. We never store the address itself, and the hash cannot be turned back into one. It is a counter and nothing else: not linked to an account, not used to identify you, not shared, and deleted after 7 days. Signing in replaces it — your allowance is then tied to your account instead.
• Service logs. To keep the service working we record each request our server handles: the time, the feature that was called, whether it succeeded, and how long it took. These are linked to your account so we can investigate a problem you report. They never include the content you submit, and are deleted after 14 days — or immediately when you delete your account. We keep anonymous totals (how often a feature was used and how fast it was) which are not linked to anyone. None of this is sent to AI providers or used for advertising.

You can delete your account and everything stored with it at any time from Settings. You can also email [email protected] to request deletion, or disconnect the app immediately from your Google account.

Changes to this policy

We may update this policy from time to time. When we do, we revise the "Last updated" date above and record which version of this policy your account has seen. If it changes, we tell you in the app rather than relying on you to come back and check.

Contact

Questions about this policy? Email us at [email protected].